A small business owner is doing several kinds of marketing at once — a website worked on, some content written, a little advertising running, a business profile kept up. Asked a simple question — is any of it working? — the owner pauses, and realises they genuinely do not know.
That not-knowing is common, and it is costly, and it is fixable. This article is the pillar of the measurement articles in this series; it sets out what it means for marketing to be working, why most small businesses measure it badly, and how a business can measure its marketing well enough to know — honestly, including about how imperfect marketing measurement always is.
A note on sources is in order, and it matters here. Peer-reviewed research is cited by author and year and listed at the end; and a substantial part of what can be said about marketing measurement is practitioner consensus — the settled common practice of the field — rather than peer-reviewed research, and that is identified plainly wherever it is so.
What “working” actually means
Before measuring whether marketing is working, a business has to be clear about what “working” means — and the clarity matters, because a vague sense of “working” is what lets bad measurement pass.
Marketing works, in the sense that matters, when it produces genuine business outcomes — customers, and the revenue they bring — worth more than the marketing cost. That is the test. Marketing that produces attention, or activity, or visible signs of effort, but not genuine customers worth more than it cost, is not working in the sense a business should care about, however busy it looks.
This is a stricter definition than the one many businesses use in practice. A business often slips into treating marketing as “working” when it sees signs of activity — visitors, clicks, followers, engagement. But activity is not outcome, and the gap between them is exactly where bad measurement lives. The whole of good measurement, as this article will show, rests on holding to the outcome definition and not drifting back to the activity one.
The reason to be strict is that the loose definition is comfortable and misleading. Activity is easy to produce and easy to see, and a business measuring activity can feel reassured while its marketing produces no genuine customers at all. The strict definition is less comfortable but it is the true one, and only the true one tells a business what it actually needs to know.
It is worth holding the strict definition in mind throughout, because the loose one will keep trying to return. Every easy, flattering activity measure a business encounters is, in effect, the loose definition inviting the business back. Good measurement is, in large part, the steady discipline of declining that invitation and keeping the question fixed on genuine outcomes.
The cost of not measuring
Before setting out how to measure well, it is worth being concrete about the cost of not measuring at all — because that cost is easy to underestimate, and underestimating it is why so many businesses never start.
A business that does not measure its marketing cannot tell its effective marketing from its wasted marketing. Some of what it does is genuinely producing customers; some is producing nothing; and without measurement the two are indistinguishable. The business therefore cannot do the obvious, valuable thing — put more into what works and less into what does not — because it does not know which is which.
The cost of this compounds quietly. A business unable to tell its effective marketing from its wasted marketing tends to keep doing both, year after year, paying the full cost of the waste indefinitely. It is not a single loss but a continuing one, and it continues precisely because the lack of measurement keeps it invisible.
There is a subtler cost too. A business that cannot measure cannot improve in a directed way; it cannot learn what its customers respond to, cannot refine its efforts toward what works, cannot get better at marketing over time. It is left improving by guesswork, if at all. The cost of not measuring is not only the waste that goes unseen; it is the improvement that never happens.
The two kinds of marketing measure
The distinction the previous section drew — activity against outcome — is the most important distinction in marketing measurement, and it deserves to be set out fully. The figure below does so.
The figure carries the article’s central caution. Activity measures are not worthless — they can be useful as early signs and diagnostic clues — but they are not the thing to judge marketing by. A business judges its marketing by the right-hand column: by whether genuine enquiries, customers, and revenue are being produced. A later article in this series treats the activity measures, and the danger of mistaking them for success, in full; this pillar’s point is that good measurement keeps its eyes on outcomes.
The measurement chain, from exposure to revenue
Activity and outcome are not unrelated; they sit at opposite ends of a chain that runs from a person first encountering a business to that person becoming a paying customer. Seeing the chain whole, as the figure below sets it out, helps a business understand what it is measuring.
The chain clarifies what the activity and outcome measures actually are: they are measures taken at different points of the same journey. Activity measures sit near the top — exposure, visits — where the numbers are large and easy to read but far from genuine value. Outcome measures sit near the bottom — enquiries, customers, revenue — where the numbers are smaller and harder to track but close to what matters. Good measurement does not ignore the top of the chain, but it judges by the bottom.
The chain also explains why activity measures are not simply to be discarded. Each stage feeds the next, so a measure near the top can be a useful early warning — a sharp fall in visits will, in time, show up as a fall in customers. The point is not that activity measures are worthless but that they are leading indicators, not verdicts: useful as early signals, misleading as final judgements.
Why most small businesses measure badly
If measurement matters, it is worth asking why so many small businesses do it badly — because the reasons, once seen, are largely avoidable.
The commonest reason is that they do not measure at all. Marketing is done — a website, some content, a little advertising — and no one ever genuinely checks whether it produced anything. The business operates on a vague feeling that the marketing is probably helping, and never tests the feeling. A business in this position cannot tell its effective marketing from its wasted marketing, and so cannot improve either.
The second reason is measuring the wrong things. A business that does measure often measures the activity column of the figure — visits, clicks, followers — because those are the numbers the tools put in front of it most readily. It ends up with a sense of being measured, and of being busy, without ever learning whether genuine customers resulted.
The third reason is the opposite failure: drowning in data. The platforms a business uses generate vast quantities of numbers, and a business that tries to attend to all of them ends up attending to none of them usefully — overwhelmed, unable to tell the few measures that matter from the many that do not. Measuring nothing, measuring the wrong things, and drowning in everything are three different failures, and good measurement, as the rest of this article describes, avoids all three.
It is worth noticing that these three failures share a single root: the absence of a clear, prior decision about what to measure and why. A business that has decided what a genuine result is, and that it will track that and little else, is protected from all three at once — it does not measure nothing, does not drift to the wrong things, and does not drown, because the decision has told it exactly where to look.
The problem of attribution
One genuine difficulty deserves an honest section of its own, because it defeats a great deal of marketing measurement: the problem of attribution — knowing which marketing produced a given customer.
The difficulty is real. A customer who finally makes contact may have encountered the business several times first — found it once in a search, seen it later in an ad, been reminded of it by a mention, and only then enquired. Asked which piece of marketing “produced” that customer, a business often cannot honestly say; the customer was produced by the combination, over time, and the journey is rarely fully visible.
The wrong response to this difficulty is to pursue perfect attribution — to try to trace every customer to an exact source. That pursuit consumes effort a small business does not have and rarely succeeds, because the journeys genuinely are tangled and partly invisible. A business that makes perfect attribution its goal will spend heavily and still not have it.
The honest response is to accept that attribution is imperfect and to work with rough, useful signals instead: broad patterns over time, the simple practice of asking customers how they found the business, the general movement of outcomes as marketing efforts change. This is practitioner consensus rather than precise science, but it is sound: a business does not need perfect attribution to measure well; it needs enough signal to make better decisions, and that is achievable where perfection is not.
There is a freeing consequence in accepting this. A business released from the impossible goal of perfect attribution is released from a great deal of wasted effort and frustration — and is free to put that energy into the rough, sufficient measurement that genuinely informs decisions. Good enough, here, is not a reluctant compromise; it is the sensible target, and it is reachable.
Channels easy to measure, and channels that are not
A further honest point: the different things a business does to be found are not equally easy to measure, and a business should understand the difference rather than be misled by it.
Some marketing is easy to measure. Paid advertising, as the advertising pillar noted, is recorded in detail by the platforms — spend, clicks, and, if set up well, the customers that resulted can be tracked with real precision. A business can see, fairly clearly, what its advertising did.
Other marketing is genuinely hard to measure. The slow effect of content marketing, the organic visibility built over years, the word of mouth that good work generates — these produce customers through paths that are diffuse, delayed, and often invisible. A customer who read a business’s article a year ago, remembered the business, and enquired today is real, but the article’s part in it may leave no trace a business can see.
The danger in this difference is subtle and serious: a business can come to value the easily measured channels over the hard-to-measure ones simply because it can see their results — mistaking measurability for effectiveness. The hard-to-measure channels are not less effective for being hard to measure; the content marketing pillar argued they are often the more valuable. A business should hold this clearly, because the next section turns it into a principle.
It is worth being concrete about how this misjudgement plays out. A business comparing its advertising, which it can measure, with its content marketing, which it largely cannot, sees clear results from the one and unclear results from the other — and concludes, naturally but wrongly, that the advertising is the better investment. The conclusion does not follow: it reflects a difference in visibility, not a difference in value.
Don’t let the measurable crowd out the important
The danger just named is important enough to state as a principle in its own right: a business must not let what is easy to measure crowd out what is genuinely important.
The mechanism by which this happens is quiet and natural. What a business can measure, it can see succeeding; what it cannot measure, it cannot see succeeding, even when it is. Over time, a business attending to its measurements drifts toward the channels that show visible results and away from the channels that do not — and if the hard-to-measure channels are genuinely valuable, as they often are, the business has been led by its measurement into a worse position.
This is the deeper reason measurement must be done thoughtfully rather than mechanically. Measurement is a servant of good decisions, and a servant that quietly biases every decision toward the measurable is a poor one. A business has to correct for the bias deliberately: to remember that the absence of a clear measurement is not evidence of absence of value, and to keep investing in genuinely valuable channels even when their value does not show up cleanly in the numbers.
The principle, then, is to measure what matters as well as it can be measured, and to keep believing in what matters even where it cannot be measured well. A business that measures only the measurable, and trusts only what it measures, will be steadily and invisibly misled. Good measurement informs decisions; it does not get to make them by default.
Asking customers how they found you
Amid the difficulty of attribution and the unevenness of what can be measured, there is one method that is simple, cheap, and genuinely informative, and it deserves its own section: asking customers, directly, how they found the business.
This is the most underused measurement method available to a small business. A business is in direct contact with its customers; it can simply ask each new one, in conversation or with a brief question, how they came to find it. The answers are imperfect — people misremember, and a customer who found a business through several touches may name only the last — but they are genuine evidence, gathered at no cost, about what is actually bringing customers in.
Over time, the answers accumulate into something valuable: a rough but real picture of which channels customers actually attribute their discovery of the business to. That picture cuts through a good deal of the attribution problem, because it comes from the customers themselves rather than from trying to reconstruct their journeys from data.
The practical recommendation is therefore plain: a business should make asking how customers found it a routine habit, and should keep a simple record of the answers. It is the closest thing a small business has to a direct measurement of its marketing’s outcomes, and it costs nothing but the asking.
This method has, moreover, a quality the data-based methods lack: it grows more valuable precisely as detailed digital tracking grows harder. As the platforms record less and attribution from data becomes more difficult, the direct evidence of customers telling a business how they found it becomes one of the soundest measurement methods a business has. It is simple, durable, and not at the mercy of changing tracking technology.
Deciding what a genuine result looks like
Good measurement begins not with tools but with a decision: a business must decide, in advance, what a genuine result of its marketing looks like.
This decision is the foundation, and it is often skipped. A business that has not decided what outcome it is looking for will measure whatever its tools happen to show — which, as the earlier section warned, is usually the activity column. A business that has decided — that has said, plainly, “a genuine result is an enquiry, a booking, a sale, a new customer” — knows what to look for and can tell whether it is getting it.
What counts as a genuine result varies by business, and each should define its own. For one business it is a phone enquiry; for another, a form submission; for another, a booking or a purchase. What matters is not which outcome a business chooses but that it chooses — that it names, concretely, the thing its marketing exists to produce, so that measurement has a definite target.
This decision also disciplines everything that follows. Once a business has defined a genuine result, the question of measurement becomes focused and answerable: is the marketing producing that result, and is the result worth more than the marketing cost? Without the decision, measurement has no anchor; with it, measurement has a clear and honest job.
This decision is also worth making with some breadth, because a genuine result is not always a sale. For some businesses an enquiry, a quote request, or a booking is the right thing to measure, since it is the point at which marketing’s job ends and the business’s own follow-up begins. A business should define the result that genuinely marks its marketing as having worked — which is the outcome marketing can be held responsible for, not necessarily the final sale.
Setting up the minimum measurement you need
With a genuine result defined, a business can set up the measurement it needs — and the guiding principle is minimum, not maximum.
The instinct, faced with measurement, can be to track everything — to assemble an elaborate dashboard of every available number. This is the drowning-in-data failure named earlier, and it should be resisted. A small business does not need comprehensive measurement; it needs enough measurement to answer the genuine questions, and no more, because measurement it cannot attend to is measurement that does nothing.
The minimum, for most small businesses, is modest. It is some way of seeing how many genuine results — as the business has defined them — the marketing is producing; some way of seeing, roughly, where they come from, including the simple habit of asking customers; and some sense of what the marketing cost against what those results were worth. This is far less than the platforms invite a business to track, and it is enough.
The practical task, then, is not to build something elaborate but to put in place the few genuine measures that matter and to ignore, deliberately, the many that do not. A business with a small, well-chosen set of outcome measures it actually looks at is measuring better than a business with a vast dashboard it cannot read.
Keeping measurement light enough to sustain
One practical point underlies whether a business’s measurement survives at all, and it deserves stating plainly: measurement has to be light enough that the business will actually keep doing it.
Measurement is itself a cost — a claim on the owner’s scarce time. A measurement system that is elaborate, time-consuming, or unpleasant to maintain shares the fate of the abandoned blog: it is kept up briefly, with good intentions, and then quietly dropped. Elaborate measurement that is abandoned tells a business less than modest measurement that is sustained.
This is the practical case for the minimum the previous section described. A small, light set of genuine measures — a few outcome figures, a simple record of how customers say they found the business — is something an owner can realistically keep up alongside running the business. It is better, by far, to have a modest measurement habit that lasts than an ambitious one that collapses.
So a business designing its measurement should ask, of the whole arrangement, a sustainability question: can I genuinely see myself keeping this up, month after month, busy or not? If the honest answer is no, the measurement is too heavy, and it should be cut back to something the business will actually sustain. Measurement that is not sustained is, in the end, measurement that does not happen.
Reviewing: the trend, not the snapshot
Having set up the minimum measurement, a business has to use it — and how it reads what it sees matters as much as what it measures.
The key principle is to read the trend, not the snapshot. A single month’s figures, looked at alone, are noisy and easy to misread: a good month may be luck, a poor one may be a seasonal dip. What genuinely tells a business something is the direction over time — whether, across many months, genuine results are growing, holding, or declining.
This is especially true for the slow-compounding channels. Content marketing and organic visibility, as their pillar articles argued, deliver their value gradually; judged by a snapshot they will almost always look disappointing, and judged by a multi-year trend they may look genuinely strong. A business that reviews by snapshot will misjudge exactly the channels whose value takes longest to appear.
The practical recommendation is to review marketing measurement periodically but not obsessively — every few months rather than every few days — and to look, each time, at the movement since last time and over the longer run. Measurement read as a trend informs genuine decisions; measurement read as a constant series of snapshots produces mostly anxiety and misjudgement.
Reading the trend also guards against a particular overreaction. A business watching its figures too closely is tempted to change its marketing in response to every short-term movement — abandoning something after one poor month, chasing something after one good one. The trend, read over a longer span, smooths that noise and reveals the genuine direction, which is the only thing a sound change of course should be based on.
Measuring the slow-compounding channels fairly
The slow-compounding channels — content marketing, organic search visibility — deserve a section of their own, because they are the channels measurement most easily misjudges, and misjudging them is genuinely costly.
The difficulty is structural. These channels deliver their value gradually and indirectly, as their pillar articles described; a customer they produce may have been influenced months ago, through a path that leaves little trace. Measured by the standards that suit advertising — fast, attributable, immediate — the slow channels will almost always look like failures, because they are being judged by a clock that does not fit them.
Measuring them fairly means measuring them on their own terms. It means reading them over long periods — a year or more, not a month — and watching the slow trend rather than the immediate return. It means using the rough signals available: whether organic visits grow over time, whether the questions customers ask are increasingly ones the content has answered, whether customers, asked how they found the business, increasingly mention its content. None of these is precise, but together they give a fair picture where a precise, fast measure would give a false one.
The honest principle is that a channel should be measured by a standard that matches how it works. To hold the slow channels to advertising’s standard of fast, clean attribution is not rigour; it is a category error that will lead a business to conclude its most valuable work is failing. The slow channels are measured fairly only when they are measured slowly.
For a small business this has a direct, practical consequence. When the time comes to review marketing and decide where to put effort, the slow channels should be given the long view they require — judged on a year or more of trend — rather than being quietly cut because a recent snapshot looked thin. A business that judges its slow channels by a fast measure will tend to dismantle exactly the work that was building its most durable value.
What measurement is for
It is worth stating plainly, before the practical conclusion, what all this measurement is actually for — because a business that loses sight of the purpose tends to measure badly.
Measurement is not for producing numbers. It is not for assembling reports, or for the reassurance of seeing figures, or for being able to say the business measures things. Measurement that produces numbers no one acts on has done nothing; the numbers are not the point.
Measurement is for informing decisions. Its entire purpose is to help a business decide better: to keep doing what is working, stop or fix what is not, and put its limited time and money where they produce the most genuine result. A measurement that could not change a decision is a measurement not worth taking; a measurement that does change one has done its job.
This purpose is the test for everything in this article. A business deciding what to measure, how much, and how often should ask, of each choice: will this help me make a better decision? Measurement that passes that test is worth the effort; measurement that does not — however precise, however busy it makes the business feel — is not. Measurement serves decisions, or it serves nothing.
This purpose also sets the right scale for the whole effort. Because measurement exists to inform decisions, a business needs exactly as much of it as its decisions require — no less, or it decides blindly, and no more, or it spends effort producing numbers that change nothing. The decisions a business actually faces are the measure of how much measurement is enough.
Holding the purpose clearly also rescues a business from a quiet discouragement. Measurement can feel like an obligation — one more thing a business is told it must do. Seen as the service of decisions, it is not an obligation but a tool the business uses for its own ends: a way of spending its limited time and money where they do the most good. Measurement is not a duty owed to anyone; it is how a business looks after itself.
A practical approach to measuring your marketing
The article’s argument resolves into a practical approach, and the table below sets out, for the main channels, the easy measure to be wary of and the measure that genuinely matters.
| Channel | An easy measure to be wary of | The measure that genuinely matters |
|---|---|---|
| Search and SEO | Visitor numbers alone | Enquiries and customers that visits produced |
| Content marketing | Page views and time on page | Whether content draws people who become customers |
| Paid advertising | Impressions and clicks | Customers won, against what the advertising cost |
| Local SEO | Profile views | Calls, direction requests, and visits that resulted |
| Social media | Followers, likes, shares | Genuine enquiries or customers it brought |
The approach, in short, is this: define what a genuine result looks like for the business; set up the minimum measurement needed to see those results and roughly where they come from; make a habit of asking customers how they found you; read the outcomes as a trend over months, not as snapshots; do not let the easily measured crowd out the genuinely important; and treat every measurement as a servant of a decision. A small business that measures this way will be able to answer the question it could not before — is the marketing working? — honestly, and act on the answer.
Concluding remarks
Measuring whether marketing is working starts with being clear about what “working” means: marketing works when it produces genuine business outcomes — customers and revenue — worth more than it cost, not when it merely produces activity that looks like success.
That distinction, between activity measures and outcome measures, is the most important in the subject. The two sit at opposite ends of a chain that runs from exposure through visits and engagement to enquiries, customers, and revenue; good measurement does not ignore the top of the chain but judges by the bottom. Most small businesses measure badly — not measuring at all, measuring the wrong things, or drowning in data — and all three failures are avoidable.
Measurement is genuinely imperfect: attribution is hard, and a business should work with rough signals rather than chasing a perfect tracing of every customer; channels differ in how easily they can be measured, and a business must not let the measurable channels crowd out the genuinely important ones it cannot see as clearly. The practical path is to define what a genuine result looks like, set up the minimum measurement needed, ask customers directly how they found you, read the trend rather than the snapshot, and treat every measurement as a servant of a decision. Measurement exists to inform decisions; numbers no one acts on are numbers not worth taking.
The next articles in this series narrow from this overview: first to the handful of measures a small business should genuinely watch, and then to the vanity metrics that look good and mean nothing.
Future developments
Marketing measurement is an area where the tools change constantly and the principles do not, and a pillar article should be clear about which is which.
The tools — the analytics platforms, the dashboards, the tracking mechanisms — change continually, and some current measurement practices will be done differently in a few years. There is also a genuine current of change worth naming: as privacy norms and rules tighten, some of the detailed tracking that measurement has relied on is becoming less available, which makes precise attribution harder rather than easier. A business should expect the mechanics of measurement to keep shifting.
The principles, though, do not depend on the tools. The distinction between activity and outcome, the imperfection of attribution, the danger of letting the measurable crowd out the important, the discipline of measuring against a defined genuine result, the reading of trends over snapshots, the purpose of measurement as the service of decisions — none of these is a feature of a particular platform. They follow from what measurement is for, and they will hold whatever the tools become.
For a small business the steady conclusion is to invest in understanding the principles rather than in mastering any current tool. A business that grasps what it should measure, and why, can apply that understanding through whatever tools the coming years bring — and as detailed tracking grows harder, the simple, durable methods this article has stressed, above all the habit of asking customers directly, become more valuable, not less.
Related reading
- Small business marketing in 2026: a complete guide
- How to prioritise marketing channels on a limited budget
- The handful of metrics a small business should watch
- Vanity metrics: numbers that look good and mean nothing
- Why visitors leave your site without contacting you
References
Broder, A. (2002). A taxonomy of web search. ACM SIGIR Forum, 36(2), 3–10.
Nelson, P. (1970). Information and consumer behavior. Journal of Political Economy, 78(2), 311–329.
Stigler, G. J. (1961). The economics of information. Journal of Political Economy, 69(3), 213–225.
